To provide a constant supply of "universal blood type" group O red blood cells (RBC), we are attempting to convert groups A, B and AB RBC to group O by enzymatic methods. We have already demonstrated the safety and efficacy of multi-unit and repeat transfusions of group B O converted RBC, and will now begin in vivo studies in which we have removed the A epitopes from group A RBC to determine whether these cells will be nonimmunogenic when they are transfused to recipients of any blood group. Such enzymatic conversion is effected using an exoglycosidase, alpha-N-acetylgalactosaminidase ("A-zyme"), with/without an endoglycosidase, endo-beta-galactosidase ("endo"). Due to the number and complexity of multiple A epitopes on the group A RBC, it is not yet known whether internal as well as external A epitopes, (internal A epitopes are present on some blood group A carbohydrate chains) must be removed to achieve efficient conversion. Our early small-volume in vivo infusions to normal volunteers (feasibility studies) will demonstrate whether one or both enzymes, or some other combination of relevant enzymes, will be necessary to remove sufficient A epitopes to produce nonimmunogenic cells with sufficient in vivo lifespan (as assessed with chromium-51 label). Pending these results, we will move to larger volume transfusions to group B and O volunteers (conducted under an IND) in which we will further assess the safety, as well as the efficacy, of converted A O RBC. These studies will include a comprehensive assessment of the effects of large-volume transfusions, including extensive hematology, clinical chemistry, urine and serologic analyses, all of which are capable of detecting subtle transfusion reactions and demonstrating product efficacy. Finally, patient subjects will be enrolled.